Student ability to learn at home – An introductory look at student access to remote learning resources

1SiyaphambiliWe Are Moving Forward

Student ability to learn at home| An introductory look at student access to remote learning resources |

Emma Whitelaw

Samantha Culligan

Nicola Branson

2

STUDENT ABILITY TO LEARN AT HOME: AN INTRODUCTORY LOOK AT STUDENT ACCESS TO REMOTE LEARNING RESOURCES

Emma Whitelaw, Samantha Culligan and Nicola Branson1

1 Siyaphambili post-school research group in SALDRU, order of listed authors names randomly assigned.

2 Some institutions have not yet had students return, but plan to reintegrate students by the end of August. Other institutions (e.g. Walter Sisulu University) have had a delayed response to the implementation of remote learning.

3 We do not restrict the analysis to households containing students as the sample size is too small. This means our aggregated municipality estimates will have greater variance. Instead, we assume that households containing students and households without students in a given municipality are similar in regard to our variables of interest.

4 Public university data is from 2018 and TVET college data is from 2019. Although this data is not from 2020, we do not have reason to believe the demographics of the student body have shifted significantly in the last two years.

The rise of the Covid-19 pandemic led South African

president, Cyril Ramaphosa, to declare a national

state of disaster on 15 March 2020. In response to

the announcement, tertiary education institutions

and student residences around the country closed

in order to mitigate the spread of COVID-19. These

institutions were required to come up with remote

teaching and learning solutions in a relatively short

period of time. While virtual classes on personal

tablets may have become the global norm, many

South African students lack access to internet

and data connectivity, and may rely on shared

or mobile devices off which to learn. In addition,

glaring structural inequalities plague a multitude

of socio-economic factors in South Africa. These

factors shape the household environment in

which many students have found themselves, and

in which they have been expected to learn new

academic material.

Household access to electricity, a stable internet

connection and a suitable device dictate to a

large degree the quality of students’ remote

learning – if it is taking place at all. Although

institutions and residences have implemented a

phased return of the most vulnerable students,

those who have not yet been able to return face

continued constraints to their learning.2 The

impact of protracted online learning on those

returning is unlikely to be negligible either.

Therefore, existing household inequalities are

likely to have disadvantaged, and continue to

disadvantage, students in this time.

In this review, we provide high-level descriptive

insights into the characteristics of home

environments for students from public universities

and Technical and Vocational Education and

Training [TVET] colleges in South Africa. Using

the 2016 Community Survey, we construct

indicators for access to electricity, access to

a stable internet connection and access to a

device (tablet or computer) in the home. We

define these three resources as the minimum

requirements for remote learning to take place.

Each indicator is aggregated to the municipality

level to estimate the proportion of all households3

in each municipality with access to the respective

indicators. A single indicator which measures if a

household has access to all three of these resources

is also created and aggregated to the municipality

level. This indicator is referred to as the ‘remote

learning’ indicator. We then map students’ home

postal code, collected in institutional4 databases,

to the municipality-level indicators created using

the 2016 Community Survey.

Student ability to learn at home – An introductory look at student access to remote learning resources

3

It should be noted that these indicators are crude

in two ways. Firstly, we can only elicit the average

characteristics of households in students’

municipalities, and not the characteristics of

student households themselves. Secondly, the

Community Survey data does not address access

to devices or the internet in a very detailed way,

5 The share of students is calculated as the number of students in a municipality over the total population of students.

6 Municipalities are ranked based on this share and grouped such that the sum of the shares approximately equals 25% in each coloured category. One of the 213 municipalities in South Africa has no resident students from a university or TVET college. This is Maphumulo municipality in KwaZulu-Natal.

which hinders our ability to comprehensively

unpack the factors affecting student remote

learning. Nonetheless, the results are broadly

illustrative of low levels of access to the necessary

(although not sufficient) remote learning

‘resources’ in municipalities in which students

reside.

Distribution of students

To contextualise the analysis, we first assess

the distribution of university and TVET

college students nationally. On the whole, the

provincial distribution of students mirrors overall

population density closely; provinces with low

population density have low student numbers

(Statistics South Africa [StatsSA]). Figure 1 shows

the distribution of university and TVET students

across municipalities5. The figure shows that

university student origins are more concentrated

than TVET student origins. For example, just

over 25% of all students in the university sector

come from just three municipalities; 11.55% from

the City of Tshwane, 9.48% from eThekwini, and

6.25% from Ekurhuleni. The next 25% of students

reside across 8 municipalities, and the remaining

50% of university students are spread over the

remaining 206 municipalities.6

For TVET students the concentration is lower; 25%

of all students in the TVET sector come from six

municipalities. The municipality with the highest

share of TVET students is the City of Cape Town

(6.89% of TVET students), followed by those most

densely populated with university students too,

namely: Ekurhuleni (5.6%), eThekwini (5.35%) and

the City of Tshwane (4.99%). The following 25%

of TVET students come from 12 municipalities in

total, with the remaining 50% distributed across

194 municipalities. The average characteristics

of households in these more densely populated

municipalities will contribute more significantly

to the aggregate pictures of student connectivity

that follow. Results should be understood with

these distributions in mind.

4

Figure 1: The distribution of university and TVET college students by municipality

Notes to Figure 1:

Source: Authors’ own calculations using data from Higher Education Management Information System [HEMIS] 2018,

Technical and Vocational Education and Training Management Information System [TVETMIS] 2019.

Sample: Number of students: N=1 043 646 (University); N=648 498 (TVET).

Student ability to learn at home – An introductory look at student access to remote learning resources

5

Indicator #1: Electricity access

7 Data are weighted using household-level post-stratification weights.

Levels of electricity access in South Africa are

good. Statistics South Africa [StatsSA] reports

that 85.4% of households had access to electricity

in 2013. Using the 2016 Community Survey

data, we estimate that 87.55% of households

have access to prepaid or metered electricity.7

It is thus not unexpected that average level of

access to electricity in municipalities in which

students reside is high, 85.96%. For this reason,

we do not report a figure alongside this indicator.

The majority (70%) of university students reside

in municipalities where more than 85% of

households have access to electricity, as do

the majority of TVET students (69,76%). TVET

students fare slightly worse on average, but only

marginally so. As noted, however, we cannot infer

actual access levels in student households.

Indicator #2: Internet access

Next, we look at internet access. Fewer than

30% of households in all municipalities in South

Africa have internet access. The municipality with

the highest share of households with access to

internet services is Richtersveld municipality in

the Northern Cape, at 28.05%. Figure 2 shows

the proportion of students living in municipalities

where fewer than 10%, between 10% and 20%,

and more than 20% (with an upper bound of

28.05 %) of households have access to an internet

connection respectively.

The majority (53.15%) of university students

come from municipalities in which 10-20% of

households have access to the internet. On

the other hand, the majority of TVET students

(48.9%) reside in municipalities in which fewer

than 10% of the households have access to an

internet connection. TVET students, in addition

to facing more severe institutional barriers to

online learning (e.g. lack of e-learning platforms),

appear more likely than university students to

face barriers to online learning through lack of

connectivity in their homes.

6

Figure 2: Percentage of students by share of households in their home municipality with internet access

38,9%

53,2%

7,9%

University students

48,9%

42,4%

8,7%

Share of students residing in municipalities in which fewer than 10% of households have a working internet connection

Share of students residing in municipalities in which 10%-20% of households have a working internet connection

Share of students residing in municipalities in which 20%-30% of households have a working internet connection

TVET students

Notes to Figure 2:

Source: Authors’ own calculations using data from Community Survey 2016 (StatsSA), HEMIS 2018, TVETMIS 2019. Com-

munity survey data are weighted using household-level post-stratification weights.

Sample: N=1 043 646 (University); N=648 498 (TVET).

Low levels of connectivity are supported by

results of a recent survey of South African youth

(ages 18 to 35) by the Southern Africa Labour

and Development Research Unit [SALDRU] and

UNICEF South Africa. The survey finds that 11%

of respondents have access to the internet, and

only 17% said they have access to a data bundle.

In an attempt to remedy this dire connectivity

situation, some institutions are offering data

bundles to students, and certain TVET colleges

are broadcasting material on radio platforms.

However, even in this regard, where students face

network connectivity or signal issues, students will

be no better off for having data (see this article

on the digital learning divide for students’ stories

regarding connectivity issues).

Indicator #3: Device access

Figure 3 shows the proportion of students

living in municipalities where the average

number of households with access to a device

(computer or tablet, which may or may not be

shared) is less than 15%, between 15% and 30%

and greater than 30% (with an upper bound

of 46.46%) respectively. The highest share of

households with access to a device is 46.46%

in the Gamagara municipality in the Northern

Cape, closely followed by Midvaal municipality

in Gauteng where 46.29% of households have

access to a device. Figure 3 shows that on

average, just under half of university students

reside in municipalities where between 30% and

47% of households have access to a device. In

comparison, only one third of TVET students

reside in municipalities with similar device access

levels. The fact that the remaining two thirds of

TVET students reside in municipalities where

fewer than a third of households have a device

is staggering. Although some universities have

provided vulnerable students with laptops, the

distribution of laptops by some institutions has

been delayed.

Student ability to learn at home – An introductory look at student access to remote learning resources

7

Figure 3: Percentage of students by share of households in their home municipality with access to a

computer or tablet in the home

University students TVET students

12,6%

42,2%

45,1%

Share of students residing in municipalities in which fewer than 15% of households have a computer or tablet in the home

Share of students residing in municipalities in which between 15% and 30% of households have a computer or tablet in the home

Share of students residing in municipalities in which between 30% and 47% of households have a computer or tablet in the home

16%

51,1%

32,3%

Notes to Figure 3:

Source: Authors’ own calculations using data from Community Survey 2016 (StatsSA), HEMIS 2018, TVETMIS 2019. Com-

munity survey data are weighted using household-level post-stratification weights.

Sample: N=1 043 646 (University); N=648 498 (TVET).

Combined remote learning indicator

Lastly, Figure 4 combines our three summary

measures into one ‘remote learning’ indicator.

This remote learning indicator captures the share

of households in a municipality that have access

to all three resource types, namely: electricity

access, internet access and access to a device

(computer or tablet). Given that the share of

households in a given municipality with access

to a stable internet connection is particularly

low, our ‘remote learning’ indicator is bounded

above by our internet indicator. Generally, a

household would need electricity and a device

to successfully use the internet. It is also unlikely

that households will have access to the internet

but not have access to electricity and a device.

More than 50% of both university and TVET

students reside in municipalities where fewer

than 10% of households have access to all three

resources. Thirty-nine percent of TVET students

come from municipalities where fewer than

4.4% of households have access, compared to

29% of university students. This means a higher

proportion of TVET students compared to

university students are clustered at the lower end

of the distribution, once again foregrounding the

relative disadvantage of TVET students regarding

ability to learn remotely.

8

Figure 4: Percentage of students by share of households in their home municipality with access to

electricity, internet and a computer or laptop in their home (access to all remote learning resources)

0,00%

5,00%

10,00%

15,00%

20,00%

25,00%

30,00%

35,00%

40,00%

45,00%

0% - 4.40% 4.41% - 8.80% 8.81% - 13.20% 13.21% - 17.60% 17.61% - 22.00%

University TVET

Notes to Figure 4:

Source: Authors’ own calculations using data from Community Survey 2016 (StatsSA), HEMIS 2018, TVETMIS 2019. Com-

munity survey data are weighted using household-level post-stratification weights.

Sample: N=1 043 646 (University); N=648 498 (TVET).

At the time that South Africa’s nationwide

lockdown was announced, the closing of

educational institutions and the uptake of

remote learning was intended to be a relatively

short-term solution to continuing education

under lockdown circumstances. However, the

Department of Higher Education and Training

plans to continue the implementation of remote

learning across the system (mixed with contact

learning where possible), with the academic year

expected to extend into the early part of 2021.

From this preliminary analysis, however, it is

clear that students at both university and TVET

colleges lack adequate resources for remote

learning, especially for such a prolonged period

of time.

Although solutions to remedy the situation have

been implemented to varying degrees across

the sector, these have been implemented in

an uncoordinated manner, and students have

remained underserved. It seems, additionally,

that judging solely on the average municipality-

level characteristics of students’ homes, TVET

students fare worse than university students.

However, we acknowledge that varying degrees of

institutional advantage exist across the university

(and TVET college) sector too. This analysis does

not capture the nuance in differential privilege -

or lack thereof. We leave addressing this salient

feature of the sector to future work.

Student ability to learn at home – An introductory look at student access to remote learning resources

9

ReferencesDe Lannoy, A. & Mudiriza, G. 2020. The state of youth wellbeing in South Africa during the Covid-19 pandemic.

Available: https://www.saldru.uct.ac.za/wp-content/uploads/U-report_survey-factsheet-2July2020.

pdf?fbclid=IwAR0qRKMXx1LHfJKdP_rjO3Bds1RmCToQZnx6ml39LGl7zwg3yjkSW9vB7oA.

Mafolo, K. 2020. The poor at universities left behind in lockdown, says student union. Daily Maverick. 25 June. Available:

https://www.dailymaverick.co.za/article/2020-06-25-the-poor-at-universities-left-behind-in-lockdown-says-student-

union/#gsc.tab=0 [2020, 18 August].

Phased-in return of students to tertiary institutions. 2020. SABC Digital News. 11 July. Available: https://www.sabcnews.

com/sabcnews/phased-in-return-of-students-to-tertiary-institutions/ [2020, 18 August].

Ngqakamba, S. 2020. 2021 university academic programme likely to start be later than usual – Nzimande. News24. 8 July.

Available: https://www.news24.com/news24/southafrica/news/2021-intake-of-first-year-university-students-will-

likely-be-later-than-usual-nzimande-20200708 [2020, 18 August].

Macupe, B. 2020. Walter Sisulu University is playing catch-up with online learning. Mail & Gaurdian. 28 July. Available:

https://mg.co.za/education/2020-07-28-walter-sisulu-university-is-playing-catch-up-with-online-learning/ [2020, 30

July].

Statistics South Africa [StatsSA]. N.d. Statistics by Place. Available: http://www.statssa.gov.za/?page_id=964 [2020, 20

August].

Statistics South Africa [StatsSA]. N.d. Household Service Delivery Statistics. Available: http://www.statssa.gov.za/?page_

id=739&id=2 [2020, 20 August].

Saving the 2020 academic year. 2020. Parliamentary Monitoring Group. 24 June. Available: https://pmg.org.za/

committee-meeting/30530/ [2020, 20 August].

Bangani, Z. 2020. The digital learning divide. New Frame. 18 June. Available: https://www.newframe.com/the-digital-

learning-divide/ [2020, 18 August].

DataHigher Education Management Information System [HEMIS]. HEMIS 2018 [dataset]. Department of Higher Education

and Training, 2020. Pretoria.

Technical and Vocational Education and Training Management Information System [TVETMIS]. TVETMIS 2019 [dataset].

Department of Higher Education and Training, 2020. Pretoria.

Statistics South Africa [StatsSA]. Community Survey 2016 [dataset]. Version 1.0.0 Pretoria: Statistics South Africa

[producer], 2017. Cape Town: DataFirst [distributor], 2017.

10

Acknowledgments

Support from the Kresge and Spencer

foundations for this research activity

is hereby acknowledged. Opinions

expressed, and conclusions arrived at,

are those of the authors and cannot

necessarily be attributed to these

foundations.